|Home | About | Journals | Submit | Contact Us | Français|
The gene encoding resistance to methicillin and other β-lactam antibiotics in staphylococci, mecA, is carried on a genomic island, SCCmec (for staphylococcal cassette chromosome mec). The chromosomal excision and integration of types I to IV SCCmec are catalyzed by the site-specific recombinases CcrA and CcrB, the genes for which are encoded on each element. We sought to identify the relative contributions of CcrA and CcrB in the excision and integration of SCCmec. Purified CcrB but not CcrA was shown to mediate the gel shift of chromosomal target integration sequences (attB) in electrophoretic mobility shift assays. However, preincubation of CcrB-DNA complexes with increasing concentrations of CcrA blocked gel shift. The interaction of CcrB and CcrA was confirmed by Escherichia coli two-hybrid analysis. SCCmec excision mediated by plasmid-encoded and inducible ccrA, ccrB, or both genes was assessed by PCR in Staphylococcus aureus. CcrB alone could mediate excision but excision was at an alternate att site (attR2) within the right extremity of SCCmec. In contrast, both CcrB and CcrA were required to mediate excision at the chromosomal attB site (called attR when SCCmec is integrated). Insertion of a plasmid containing the SCCmec att site (attS) into the chromosome required both CcrA and CcrB, but CcrA overexpression lowered integration frequency. Thus, while CcrB binds DNA, interaction between CcrA and CcrB, in a precise ratio, is required for attB site-specific excision and SCCmec chromosomal insertion.
Staphylococcus aureus is one of the most common causes of serious human bacterial infections, both in the hospital and the community (33). Therapy of these infections is made more difficult by the development of resistance to drugs with antistaphylococcal activity such as the beta-lactam antibiotics. Resistance to beta-lactam antibiotics in staphylococci is mediated by a beta-lactamase and by a beta-lactam-resistant target transpeptidase, penicillin-binding protein 2a (PBP2a) (4, 5, 8). However, while the beta-lactamase has a narrow substrate specificity, limited to penicillins, PBP2a resists inactivation by all beta-lactam antibiotics and can cross-link peptidoglycan when all other target PBPs are rendered nonfunctional by beta-lactams. The latter is called methicillin resistance and is the most important clinical resistance phenotype among staphylococci (8) The gene for PBP2a, mecA, is located on a genomic island called SCCmec (for staphylococcal cassette chromosome mec) that is integrated into the staphylococcal chromosome at a specific site. In addition to mecA, all SCCmec elements carry intact or mutant mecA regulators (mecR1/mecI) and genes that mediate the site-specific integration and excision of SCCmec (ccr genes) (14). SCCmec elements have been typed according to the sequences of the ccr and mec complexes with five cores (types I to V) being prevalent but with considerable variation in the genetic organization within each element (14-17, 22).
SCCmec is presumed to be a mobile genetic element, which can integrate into and excise from the chromosome by site-specific recombination between a site on SCCmec (attS) and one on the chromosome (attB). attB comprises the last 15 bp of a highly conserved gene called orfX that is located near the S. aureus origin of replication (15, 19). When SCCmec is inserted, the attB sequence is duplicated at the other end of the element with the site in orfX now called attR and the one abutting the non-orfX end of SCCmec designated attL. When SCCmec excises, the attB site is reconstituted in the chromosome and the two ends of the element come together to form attS within a nonreplicating circular version of SCCmec.
The site-specific recombination of SCCmec is catalyzed by its encoded ccr recombinases, CcrA and CcrB for types I to IV and CcrC for type V. CcrA and CcrB belong to a family of large serine invertase and resolvases which consist of resolvases, invertases, phage integrases, and transposases (6, 10, 29, 31). All of them contain a conserved catalytic motif and some contain DNA-binding domains at either the N or the C terminus. The catalytic domains can either function as both integrases and excisases or as only integrases that require additional proteins to mediate excision (6, 29, 30, 31).
The ccrA and ccrB genes are part of two-gene operons of 1,350 and 1646 bp in S. aureus strain N315 encoding proteins of 52.6 and 62.7 kDa, respectively. Although there is considerable variation at the amino acid level among the CcrA and CcrB proteins found in types I to IV SCCmec, plasmid-encoded CcrA and CcrB recombinases from each type can excise SCCmec from any of the others (23). However, CcrC can only excise type V SCCmec (16). There has been little examination of the role of each of these proteins in recombination or in DNA binding. In the present study we sought to define the precise roles of CcrA and CcrB in DNA binding and in the excision and integration of SCCmec in S. aureus. This is the first step in understanding the host range of SCCmec and how it may move among staphylococcal isolates in nature.
All of the strains used in the present study are listed in Table Table1.1. S. aureus strains were cultured in tryptic soy broth (TSB) or Mueller-Hinton broth (MHB; Difco, Lawrence, KS). Escherichia coli Top10 (Invitrogen, Carlsbad, CA) was used for gene cloning. E. coli Rosetta2 DE3 (pLys; Novagen, Gibbstown, NJ) was used for protein expression and purification. The antibiotics and concentrations used were as followed: ampicillin (Ap; 100 μg/ml) for selection of E. coli strains after transformation and chloramphenicol (Cml; 10 μg/ml), erythromycin (Erm; 5 μg/ml), and tetracycline (Tet; 5 μg/ml) for selection of S. aureus strains after electroporation or transduction (Sigma-Aldrich, St. Louis, MO).
CcrA and CcrB with N-terminal glutathione S-transferase (GST) fusions were produced by introducing the genes into the SmaI/XhoI site of pGEX-6P-1 (Invitrogen, Carlsbad, CA) using the primers shown in Table Table2.2. E. coli Rosetta2 (pLysS; Novagen) was used as a host for overexpression of CcrA and CcrB. Each culture was grown at 37°C in LB medium supplemented with Ap to a turbidity at 600 nm of 1.0, cooled to 25°C, and induced with 0.1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) for 5 to 6 h. Cells were pelleted by centrifugation at 6,000 rpm for 15 min and stored at −70°C until needed. The cell pellets were suspended in 1× phosphate-buffered saline (PBS; pH 7.4) containing 10 mM dithiothreitol (DTT) and 0.1 mM phenylmethylsulfonyl fluoride. After sonication using 10 20-s bursts on ice and centrifugation for 30 min at 13,000 rpm at 4°C, the supernatant was incubated with 20 ml of Sepharose 4B (Amersham, Piscataway, NJ) slurry preequilibrated in 1× PBS (pH 7.4) and softly shaken at room temperature for 1 to 1.5 h. The slurry was packed into a column and washed with 200 ml of 1× PBS (pH 7.4). The GST-tagged protein was eluted with 10 mM reduced glutathione (pH 8.0) containing 10 mM DTT and protease inhibitor cocktail and concentrated with a YM-30 Centricon (Millipore, Billerica, MA). The protein concentration was estimated by spectrometry.
Electrophoretic mobility shift assay (EMSA) was performed according to the digoxigenin (DIG) gel shift kit protocol (second generation; Roche, Indianapolis, IN). All of the DNA fragments used in this assay were PCR amplified using the primers shown in Table Table22 and then 3′ end labeled with DIG-11-ddUTP. Portions (0.8 to 1.2 ng) of the labeled DNA fragments were incubated with 0 to 2 μg of purified protein and 0.05 μg of poly-d(I-C)/μl and 0.005 μg of poly-l-lysine/μl in the binding buffer [100 mM HEPES (pH 7.6), 5 mM EDTA, 50 mM (NH4)2SO4, 5 mM DTT, Tween 20, 1% (wt/vol), 150 mM KCl] at room temperature for 20 min. The samples were then separated on a 6% retardation gel (Invitrogen, Carlsbad, CA) and transferred to a positively charged nylon membrane (Roche). DIG chemiluminescence was detected by exposure to X-ray film.
Protein-protein interaction between CcrA and CcrB was assayed by the E. coli two hybrid system (BacterioMatch II two-hybrid system; Stratagene, La Jolla, CA) as described by the manufacturer. The ccrA gene amplified from N315 was digested by BamHI and XhoI and fused to the RNA polymerase alpha subunit gene on pTRG, yielding pWA90. The ccrB gene amplified from N315 was digested by NotI and XhoI and cloned into pBT, where it was fused to the lambda cI repressor gene, generating pWA112. The plasmids were transformed separately and together into the BacterioMatch II two-hybrid system. E. coli reporter strains and cells were plated on agar containing 5 mM 3-amino-1,2,4-triazole (3-AT) to detect histidine prototrophy, and 25 μg of streptomycin/ml was used to detect transcription of the aminoglycoside modifying enzyme, aadA, conferring streptomycin resistance.
Inducible ccrAB expression in S. aureus was accomplished by constructing a series of plasmids based on the S. aureus-E. coli shuttle vector, pCN51, which has a cadmium-inducible promoter (9). First, pCN51 was modified by replacing the ermC gene (Erm resistance) with an ApaI/XhoI fragment from pCN38 (9) containing the cat gene (Cml resistance), generating pWA48. Then, ccrA4, ccrB4, or ccrA4B4 genes from MW2 and ccrA2, ccrB2, or ccrA2B2 from N315 were amplified and inserted into the BamHI site of pWA48 to create the plasmids shown in Table Table1,1, using the primers shown in Table Table2.2. All plasmids were introduced by electroporation into the restriction-deficient S. aureus strain RN4220 and then transduced into other S. aureus strains using phage 80α as previously described (26, 28).
Isogenic strains with or without SCCmec were created by inducing the excision of SCCmec with CcrA and CcrB. Portions (30 μl) of overnight cultures of S. aureus strains carrying pWA48 or derived vectors were transferred into 2 ml of fresh TSB and grown at 37°C for 3 h until the optical density reached 0.6. Transcription of ccrAB was then induced by the addition of 5 μM CdCl2 with growth at 37°C for another 4 h. Excision was confirmed by PCR amplification, identifying an empty att site and a circular excised form as described in Results. All PCR fragments were cloned into the commercial vector pSC-B (Stratagene) for DNA sequencing, which was performed by the Nucleic Acids Research Facility at Virginia Commonwealth University (Richmond, VA).
The chromosomal ccrA2B2 gene contained within SCCmec in N315 was deleted by allelic replacement mutagenesis, using vector pKOR1 (3). This plasmid allows efficient cloning without the use of restriction enzymes and ligases, using the lambda recombination cassette (Invitrogen), and positive selection for secondary recombination. Two 1-kb fragments, one upstream and one downstream of ccrA2B2, were amplified by PCR, using the primers ccrAB-dN315-11, ccrAB-dN315-12, ccrAB-dN315-13, and ccrAB-dN315-14 as shown in Table Table2;2; digested by BamHI; and ligated overnight. The ligated product was mixed with the plasmid pKOR1 and the BP clonase II enzyme (Invitrogen), incubated at room temperature for 5 h, and transformed into E. coli TOP10-competent cells. The resulting plasmid, pWA88, was recovered from E. coli, introduced into RN4220 by electroporation, transduced into S. aureus N315 using phage 80α, and selected on TSB agar plates containing 10 μg of chloramphenicol/ml. A colony of N315 containing pWA88 was grown at 42°C to inhibit replication of the temperature-sensitive (TS) plasmid and colonies growing on chloramphenicol were again selected, presumably containing only chromosomally integrated sequences. Colonies that had undergone secondary recombination between the sequences flanking ccrAB, expelling the plasmid and the ccrAB genes, were selected by inducing the culture with 1 μg of anhydrotetracycline/ml. The plasmid contains an antisense construct to the essential gene secY that is driven by a tet-inducible promoter. Any colonies that still contain the integrated plasmid and have not undergone secondary recombination will not grow after Tet induction. Chromosomal deletion of ccrAB was confirmed by PCR amplification and Southern blotting (Roche).
attS-containing plasmids were constructed from pSRatt as described by Katayama et al. (19). The copy of native ccrA2B2 was removed from pSRatt by BamHI digestion and self-ligation, yielding pWA16, which carries the SCCmec att site and 1 kb of flanking DNA (attS), an antibiotic resistance marker (Tet resistance, tetL) and a TS origin of replication.
For further investigation of integration in vivo, another set of attS-containing plasmids was constructed, which carried only a 57-bp attS site. The 57-bp attS site was generated by synthesizing and annealing complementary oligonucleotides and cloning into the commercial vector pSC-B (Stratagene), producing pWA154. An 8.1-kb XbaI fragment containing a copy of native ccrA2B2, a ts origin, and the Tet resistance gene from pSR (19) was inserted at the XbaI site of pWA154, yielding pWA159. For the ccrA2 deletion from pWA159, a ccrB2 element was constructed from two fragments. The native promoter was amplified by using the primers ccrA-SR-1 and ccrA-SR-3, and the ccrB2 gene was amplified by using the primers ccrB-SR-1 and ccrB-SR-2. These two fragments were digested by KpnI and ligated. The ligated product was amplified with the primers ccrA-SR-1 and ccrB-SR-2 and cloned into the BamHI site of pYT3 (19), resulting in pWA191. A 6.8-kb XbaI fragment containing ccrB2 and its native promoter, a TS origin, and the Tet resistance gene from pWA191 were inserted into the XbaI site of pWA154, generating pWA192.
Strain RN4220 carrying either ccrAB driven by its own, native promoter (pWA159) or by a cadmium-inducible promoter (pWA46) was grown in MHB to an optical density at 600 nm of 0.5. The total RNA was then extracted by an RNeasy Protect minikit (Qiagen, Valencia, CA), and genomic DNA was removed by a DNase treatment and removal kit (Ambion, Austin, TX). The concentration of RNA was determined by spectrometry. After conversion of RNA to DNA by using reverse transcriptase, quantitative reverse transcription-PCR (RT-PCR) was performed to assess the expression of the ccrAB operon (Virginia Commonwealth University Nucleic Acids Research Facility).
In order to assess their DNA interaction, GST-tagged CcrA and CcrB proteins were incubated with DNA fragments containing the various attachment sites and flanking sequences. The fragments were designated attB274, attS456 (attS), attR358, and attL370, with the subscript number indicating the size of the fragment. The attB sequence was the 3′ end of the orfX gene from S. aureus RN450, attR and attL flank SCCmec from S. aureus N315, and attS was the PCR amplicon from SCCmec upon excision from N315. The EMSA revealed no gel shift of any fragment with GST-CcrA or GST alone, but all four fragments shifted with GST-CcrB. Site-specificity of GST-CcrB binding was demonstrated by the absence of gel shift when fragments were used that did not contain the core binding sequence. In addition, preincubation of CcrB with sequences containing the att sites but not those without the att sites could block gel shift, confirming the specificity of binding. However, as shown in Fig. 1A to D, there were fewer shifted complexes when attB274 and attL370 were incubated with GST-CcrB (maximum of two bands with 2 μg of protein) than for attS456 and attR358 (at least four shifted bands). This suggested that there may be a difference in the number of attachment sites present on each fragment. Jansen et al. have identified an additional att sequence (attR2, GAAGCTTATCATAAGT) within SCCmec, 84 bp from the orfX site (attR1, GAAGCATATCATAAAT) (18). Excision at each site would produce two different attS and flanking sequences in the excised circular SCCmec: attS1 and attS2. DNA fragments were prepared by PCR from attS1 (attS1103), attS2 (attS292), attR1 (attR1110), and attR2 (attR2104). As shown in Fig. 1E to H, all four sites bound CcrB protein and shifted complexes were seen by EMSA.
Although CcrA was unable to bind DNA by itself, we hypothesized that CcrA might interact with CcrB or with CcrB-DNA complexes. To test this hypothesis, increasing concentrations of GST-CcrA were added into the CcrB-DNA binding system, and EMSA was performed using a fixed concentration of GST-CcrB. As seen in Fig. Fig.2,2, the number of complexes and the amount of DNA shifted decreased with increasing GST-CcrA concentrations for all att targets. Incubation with the same concentration of GST alone had no effect (data not shown). CcrA inhibition of CcrB-mediated gel shift occurred whether CcrA was first incubated with CcrB alone or after CcrB had been added to DNA. These data suggest that CcrA has a direct effect on the CcrB-DNA interaction.
The BacterioMatch II two-hybrid system (Stratagene) was used to assess whether there was a protein-protein interaction between CcrA and CcrB. This system fuses the two proteins of interest to a bait protein (lambda cI repressor) and a target protein (the alpha subunit of RNA polymerase). If the two proteins interact, RNA polymerase displaces the cI repressor and allows transcription of reporter genes (his3, conferring histidine prototrophy and aadA, conferring streptomycin resistance). As shown in Fig. Fig.3,3, when the bait vector containing ccrB (pWA112) and the target vector containing ccrA (pWA90) were both introduced into the reporter strain (Stratagene), the cells grew on agar containing double selection. Cells with either plasmid alone or the empty vector did not grow on double selection agar.
Spontaneous excision of SCCmec, mediated by the ccrAB genes carried on SCCmec, has been shown to occur spontaneously but takes place at low frequency or after passage in vancomycin (17, 25). However, ccrAB overproduced from a multicopy plasmid leads to high-frequency excision (23). Therefore, in order to monitor SCCmec excision at a detectable frequency, we chose to overproduce ccr genes that were plasmid encoded from an inducible promoter. ccrA, ccrB, and ccrAB from either N315 (ccrA2B2) or MW2 (ccrA4B4) were ligated into pWA48 downstream of a cadmium-inducible promoter and introduced into methicillin-resistant S. aureus (MRSA) strains N315 (SCCmec type II), J35 (type IV), J52 (type IV), and COL (type I; see Table Table11 for strain descriptions). After 4 h of cadmium induction, total DNA was extracted for each strain and PCR products were sought using primers that would amplify the empty chromosomal att site (attB) and excised, circularized SCCmec (attS) as previously described (23). CcrA+CcrB and CcrB alone could mediate SCCmec excision in all of the strains examined. CcrA alone could not mediate excision, a finding consistent with the lack of DNA binding in vitro. In COL, while CcrA and B could mediate SCCmec excision after 4 h of cadmium induction, CcrB alone could only cause excision after 20 h of induction. Since the native ccrB gene in COL SCCmec is known to contain a premature stop codon, which is thought to produce a nonfunctional protein (23), this suggested that the native functional ccrAB genes on types II and IV SCCmec may have contributed to excision seen with CcrB alone in these strains. Thus, in order to eliminate any contribution of the native ccrAB genes carried on each SCCmec element in the excision process, these genes were deleted from strain N315 using pKOR1-mediated recombination, creating strain AW3. When the studies described above were repeated in AW3, the same results were seen when ccrAB was induced. However, CcrB alone produced a different-sized PCR amplicon from that seen with both proteins. The attB amplicon was larger and the attS amplicon smaller when CcrB alone was present compared to cells producing both CcrA and CcrB, as seen in Fig. Fig.4.4. Sequence analysis of the PCR amplicons showed that CcrB alone excised SCCmec at the attS2 site within SCCmec, while CcrA and B mediated excision at the attS1 site in orfX. The same excision result was seen when CcrB alone was overproduced in strains that had intact ccrAB SCCmec chromosomal sequences, but excision at attS2 occurred at a lower frequency than excision at attS1. This also explains the excision results in COL with CcrB alone since COL has no attS2 site within SCCmec. Presumably, either longer exposure to CcrB alone was required in the absence of attS2 or SCCmec was spontaneously excised. The excision data with strains COL and AW3 are summarized in Table Table33.
Integration of SCCmec was determined by assessing the ability of inducible CcrA, CcrB, or both on one plasmid to promote the integration of a second plasmid pWA16 carrying the SCCmec att site and 1 kb of flanking DNA (attS), an antibiotic resistance marker (Tet resistance, tetL) and a TS origin of replication (pE194; failure to replicate at 42°C). After induction of transcription of the ccr gene(s), the cells were grown at 42°C and plated on agar containing Tet. Any colonies that grew would have chromosomally integrated plasmid as a result of attS-mediated recombination. Controls with the TS plasmid without attS sequences were included in each experiment. Two S. aureus strains were used to assess integration, RN450 and N315ex, N315 with SCCmec excised at attR1. Integration was confirmed by picking several colonies from Tet agar and performing PCR with primer pairs, one of which was located within each end of the plasmid and the other of which was located in flanking chromosomal DNA. Sequence analysis was performed to confirm the presence of attR and attL within the amplicons. Both CcrA and CcrB were required for plasmid integration into the chromosome of either strain; neither CcrA or CcrB alone could mediate integration. These data are summarized in Table Table33.
We next sought to assess the relative roles of CcrA and CcrB in integration by separating the ccrA and ccrB genes, having one driven by its native promoter on the integration plasmid and the other on a separate plasmid, driven by the cadmium-inducible promoter. First, we incorporated the ccrAB gene complex with its native promoter from N315 on the integration plasmid pWA159 and noted a 100 to 130% integration frequency. That is, the same number or more colonies grew on Tet agar at 42°C as at 30°C. We next evaluated integration in the presence of increasing concentrations of CcrA. The ccrA gene on pWA159 was inactivated so that only ccrB was driven by its native promoter (pWA192); the only CcrA activity was due to production from the cadmium-inducible vector (pWA44). However, the cadmium-inducible promoter was found to be leaky in the absence of cadmium induction, producing enough CcrA to yield the same integration frequency as from pWA159 (100 to 130%). Uninduced transcription of CcrA from the cadmium-inducible promoter on pWA44 was 1.5 times the transcription of CcrA from its native promoter on pWA159, as determined by RT-PCR. Nevertheless, the effect of excess CcrA on integration could be observed, replicating the interference by CcrA of CcrB DNA binding seen in vitro. As the concentration of cadmium increased, producing more CcrA, the integration frequency decreased to 70.9% ± 6.0% (mean ± the standard deviation) with 1 μM cadmium (Fig. (Fig.5),5), an ~30% decrease in integration. These observations illustrate the extraordinary dependence of Ccr recombination activity on the stoichiometric relationship of the two proteins.
In this study we used gel shift experiments to assess the ability of CcrA and CcrB to bind specific sequences containing the various chromosomal and SCCmec att sites. It was not surprising that CcrB, but not CcrA, could bind and shift these sequences since secondary structure analysis of CcrB identified a C-terminal Ogr-Delta (zinc-finger-like) DNA-binding motif that was missing from CcrA (SMART [http://smart.embl-heidelberg.de/smart/set_mode.cgi?GENOMIC=1]). The location of an additional att site just internal to the orfX end of SCCmec was also suggested by the increased number of shifted bands seen when sequences containing both attB and this end of SCCmec were incubated with CcrB. The presence of this internal att site (attR2) was confirmed by EMSA using only this sequence. The relevance of this site was shown when SCCmec excision was assessed in vivo. Excision was shown to occur at either site, but CcrB alone mediated excision at the internal site, while both CcrA and CcrB were required to promote excision at the orfX att site (attR1). This observation was further confirmed by the difficulty with which SCCmec was excised from COL, which has no attR2, by CcrB alone (Table (Table3).3). Excision at attR1 versus attR2 leaves different sequences behind in the chromosome, as shown in Fig. Fig.4C4C.
The presence of more than one excision att site in clinical S. aureus isolates has also been noted by Jansen et al. (18) and Diep et al. (12). Both groups noted the presence of both alternative excision products and different chromosomal remnants generated by the presence of multiple att sites. Some of the excised elements (SCC-like and ACME) contained neither ccr genes nor mecA but could have “hitch-hiked” into the chromosome when SCCmec integrated and then excised independently of SCCmec. Independent excision of both SCC-like and ACME was demonstrated by each of these groups. The presence of alternative att sites could also explain our previous observation (24) that methicillin susceptible S. aureus contain sequences flanking the chromosomal attB site that have previously been found flanking SCCmec in MRSA. Some of these sequences contained att sites and could have been integrated by SCCmec-encoded CcrAB and left behind when only SCCmec excised.
The requirement for both CcrA and CcrB to mediate attB-specific excision and integration was also noted in these studies, as was the interaction between the two proteins in the E. coli two-hybrid assay. Furthermore, the interference of excess CcrA with both CcrB DNA binding and SCCmec integration suggests that there is a critical stoichiometry between the two that determines their optimal activity. The requirement for both proteins for SCCmec integration has been previously demonstrated by Katayama et al. (19). It is interesting, therefore, that CcrB alone can mediate excision at alternative att sites. This suggests that sequences flanking the orfX attB site are important in the recognition, binding, and recombination mediated by the CcrA-CcrB complex. Recombination has been shown in other systems, such as lambda bacteriophage integrase, bacterial transposon Tn3 resolvase, and S. aureus resolvase Sin, to require that flanking sequences be present to bind accessory factors (1, 7, 11, 27). It has been proposed that the inverted repeats present at the extremities of types I to IV SCCmec are critical for CcrA- and CcrB-mediated excision. CcrA and CcrB cannot mediate the excision of type V SCCmec which does not contain these inverted repeats (16). The CcrC recombinase, a homolog of CcrB encoded within type V SCCmec, can mediate excision of this element without help from a CcrA-like protein. This may be a model for excision by CcrB alone at alternate att sites.
The structure of the two recombinases and their interactions are unique among serine recombinases. The catalytic domains of both CcrA and CcrB fall by phylogenetic analysis into the large serine recombinase family, while the structure of CcrB, with both an N-terminal catalytic domain and a C-terminal DNA-binding domain, places it among a different group, the phage integrases (29). In addition, there are no examples of a requirement for two serine recombinases to mediate the insertion or integration of DNA (29). Some phage, such as TP901-1 and C31, have separate Xis genes that mediate excision (6, 29, 32), whereas integration is accomplished by a single integrase protein. In contrast, CcrB requires CcrA for integration but CcrB alone can mediate excision at alternate att sites. The fact that CcrC alone can mediate integration of type V SCCmec sequences could provide clues to the differential activities of these recombinases.
The analyses described here have identified novel recombinase activities of two proteins that are required for the site-specific integration and excision of an important genomic island that is responsible for the acquisition and spread of resistance to β-lactam antibiotics among staphylococci. Further elucidation of the function of these and related recombinases may help identify the molecular mechanism for the rapid global spread of methicillin resistance.
This study was supported in part by Public Health Service grant 2RO1AI035705 from the National Institute of Allergy and Infectious Diseases (G.L.A.).
We are indebted to Teruyo Ito for providing plasmids for this study.
Published ahead of print on 9 April 2010.